We have designed and reported a highly sensitive technique to detect minimal disease in follicular lymphoma using PCR. In a prospective study, peripheral blood samples obtained from several hundred lymphoma patients are being analyzed for the presence of circulating t(14;18) phoma cells. We have undertaken this longitudinal clinical trial to determine whether relapse can be predicted early using PCR. To expand the applicability of PCR to the diagnosis of lymphoma, we are (1) developing oligonucleotide primers specific for each of the four breakpoint regions so that nearly all (approximately 90% of t(14;18) lymphomas can be detected, (2) designing PCR amplification from paraffin-embedded fixed-tissue and (3) investigating bcl-2 expression in tissue sections using our anti-bcl-2 antiserum and monoclonals to the bcl-2 protein now under development. We have shown that the T cell receptor delta chain gene is frequently rearranged in both T and pre-B ALL. By contrast, in most mature T cell neoplasms, we found that both T-delta alleles have been deleted from the genome. We are now investigating the manner whereby T-delta rearrangement and deletion occurs during T cell development by analyzing these events in our alpha/beta T cell neoplasms. To determine whether the sensitive detection of clonality by gene rearrangement analysis is useful in differential diagnosis and predicts a malignant outcome, !we have selected a specific anatomic site, the orbit, prone to be involved by Lymphocytic infiltrates whose malignant behavior is difficult to predict by morphology alone. Among 20 such patients gene rearrangement analysis proved to be more sensitive than immunophenotyping for detecting B cell clones. Clinical follow-up, rebiopsy and clonal analysis demonstrated that gene rearrangements can predict a malignant outcome.